Abstract B2: Switching to aflibercept treatment resulted in greater tumor responses than continuous bevacizumab treatment in patient-derived xenograft models of colorectal cancer.

Abstract

Abstract Background: The recombinant fusion protein aflibercept (afl) binds VEGF-A, -B, and placental growth factor (PlGF). Based on an OS benefit observed in the phase 3 VELOUR trial, aflibercept in combination with FOLFIRI was approved for patients with metastatic colorectal cancer (mCRC) following a prior oxaliplatin-containing regimen. In vitro, afl inhibited VEGFR signaling more potently and blocked cell migration more effectively vs. bevacizumab (bev) which only binds VEGF-A. Recent in vivo studies comparing the anti-tumor activity of afl vs. bev were conducted in patient-derived xenograft (PDX) CRC models. The PDX models are clinically relevant as they represent common genetic backgrounds in CRC and were derived from primary tumors or from metastases. A sequential treatment regimen of aflibercept post-bev vs. continuation of bev treatment was investigated. Methods: Forty eight different CRC adenocarcinoma tumors were engrafted subcutaneously (SC) into NMRI nude mice to generate 48 distinct PDX CRC models. ELISA and Luminex measured the expression of mouse and human VEGF-A and mouse and human PlGF in untreated tumors. Tumor-bearing mice (8 mice/group) were dosed with afl, bev, or placebo 2x/wk for 3 wks and SC tumor measurements were recorded 2x or 3x/wk. In 3 PDX models, mice that were treated with bev were randomized and either switched to afl treatment or continued to receive bev. Activity was expressed as relative change in tumor volume between treatment (ΔT) and control (ΔC) at 3 wks. A ΔT/ΔC ≤0% indicated complete tumor stasis. Student's t tests evaluated statistical differences in final tumor volumes. Results: Aflbercept induced complete tumor stasis in 31/48 PDX CRC models and bev in 2/48 PDX CRC models. The PDX models were characterized into 2 phenotypes based on statistical analysis: phenotype A, where afl activity was greater than bev activity (39 /48 PDX), and phenotype B, where there was no significant difference between the 2 drugs (9/48 PDX). Three models in which bev treatment resulted in eventual tumor escape vs. continuous inhibition by afl were selected for sequential treatment of bev followed by afl. Afl treatment began after bev treatment resulted in a greater than 2-fold increase in tumor volume. Greater anti-tumor activity occurred when afl was administered after bev vs. continuous bev treatment. At the end of the study, the tumor volumes were statistically significantly different between the afl and bev groups (T test p values < 0.0001). Analysis of angiogenic factors revealed that the tumor cells were the major source of VEGF whereas PlGF was primarily produced by the tumor stroma. Conclusions: Aflibercept induced complete tumor stasis more frequently in CRC PDX models than bevacizumab. As tumor levels of hVEGF-A were far greater than mVEGF-A, the inability of bevacizumab to bind mVEGF-A is unlikely to explain the higher and more consistent activity of aflibercept. In all 3 PDX models tested, switching to aflibercept treatment after bev treatment resulted in a greater tumor response than continuous bev treatment. The greater response to aflibercept than bev after an initial treatment with bev suggests that individual patients may derive more benefit from aflibercept than bev in second line treatment after an initial bev-based regimen. Citation Information: Mol Cancer Ther 2013;12(11 Suppl):B2. Citation Format: Marielle Chiron, Rebecca G. Bagley, Jack Pollard, Christophe Henry, Parminder Mankoo, Loic Vincent, Catherine Geslin, Tatjana Kloss, Donald A. Bergstrom. Switching to aflibercept treatment resulted in greater tumor responses than continuous bevacizumab treatment in patient-derived xenograft models of colorectal cancer. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2013 Oct 19-23; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2013;12(11 Suppl):Abstract nr B2.